IHHT – hypoxia training

What happens in the device

You lie down, breathe through a mask, and a device controls the oxygen content of the air you breathe. Normal air contains 21 percent oxygen. In the hypoxia phases it is lowered, in the published studies to 10 to 14 percent, in one paper to 12 percent. This is followed by the recovery phase with enriched air, 30 to 40 percent depending on the protocol.

The alternation runs in cycles over a session of about 40 to 50 minutes, controlled via saturation measured at the finger. In studies there were 2 to 3 sessions per week over 3 to 7 weeks, in one paper 15 sessions in 3 weeks.

The mechanism that holds up

Cells have an oxygen sensor: when oxygen becomes scarce, the transcription factor HIF stabilizes and switches on an adaptation program – more red blood cells via erythropoietin, new capillaries, altered glucose metabolism, antioxidant protective systems. For discovering how cells sense and adapt to oxygen, Kaelin, Ratcliffe and Semenza received the 2019 Nobel Prize in Medicine.

This is the same stimulus that altitude training camps have used for decades. The hyperoxia phase was added later because it makes recovery faster and the session more tolerable.

The most popular story goes further: weak mitochondria are supposed to tolerate the oxygen stress less well and be broken down, while the robust ones multiply. This hypothesis is plausible from cell and animal models, but direct proof in humans is lacking. A 2022 systematic review notes that moderate hypoxia activates telomerase and delays senescence markers in the lab – and that no study shows a longer life expectancy in humans.

Where the procedure comes from

Intermittent hypoxia is not an invention of the longevity industry. It was developed from the 1970s onward in Soviet aerospace medicine to prepare cosmonauts, pilots and mountaineers for extreme conditions, and then moved into rehabilitation medicine.

This explains the structure of the evidence: many decades of practical experience, mostly within one language area, and controlled studies to Western standards only in recent years.

The study everyone is talking about

The best-known paper concerns long COVID. 145 patients in inpatient rehabilitation received the training in addition to the rehab program or the rehab program alone. The primary endpoint was 6-minute walking distance: plus 91.7 versus plus 32.6 meters, that is, 2.8 times as much. Stair-climbing performance, shortness of breath and exhaustion also turned out better.

The design limits what can be concluded: allocation was not randomized, and the study was unblinded. With an endpoint that depends on willingness to exert oneself, this is exactly the constellation in which expectation plays a part. The paper prompted a methodological critique, together with an authors’ reply, in the same journal.

What else is available

In geriatric patients, the procedure was tested double-blind against sham treatment: 34 patients aged 64 to 92 completed a training program, one half additionally with hypoxia-hyperoxia, the other with room air. The dementia screening test improved by 16.7 percent versus minus 0.39 percent, the clock-drawing test by 10.7 versus minus 8 percent, walking distance by 24.1 versus 10.8 percent.

The same study group, however, also reported a negative result for the same patients: for mobility and self-rated health there was no difference between the groups, neither in the mobility test nor in standing up and walking nor in the activities of daily living index. Both findings must be read together.

In addition, there are small studies. A pilot study in 21 people with mild cognitive impairment (2019) was retracted by the journal in 2024 and no longer counts here. In 65 patients with metabolic syndrome, blood pressure fell markedly, along with cholesterol and a fatty liver index. And in 120 randomized bypass patients, fewer complications occurred after hypoxic-hyperoxic preconditioning: 23.3 versus 41.1 percent.

The comparison with training

Endurance and interval training produce the same adaptations – HIF activation, new mitochondria, new capillaries – plus effects on muscles, bones and mental health. The procedure is a training imitation for the oxygen sensor, without the musculoskeletal system.

This determines the order: whoever can train, trains. The device is a supplement or substitute for those for whom training is not, or not yet, possible – rehabilitation, severe exhaustion, advanced age. That is exactly where the studies are.

What users report after a course of sessions is mostly soft: better sleep, more stable daytime energy, lower resting heart rate, occasionally slight head pressure at the beginning. These are reports, not evidence, and with a relaxed lying-down appointment with personal attention, part of this is an expectation effect. All the more important are measurable values before and after: resting heart rate, exercise test, blood values.

What the studies show

Doehner et al., Journal of Cachexia, Sarcopenia and Muscle 2024

Controlled pilot study in 145 long COVID patients in inpatient rehabilitation, 74 percent women, mean age 53 years. 70 patients additionally received hypoxia-hyperoxia training 3 times per week, 75 standard rehabilitation. Primary endpoint 6-minute walking distance: plus 91.7 versus plus 32.6 meters. Allocation was not randomized and the study was unblinded; the paper prompted a published methodological critique together with an authors’ reply.

Bayer et al., Alzheimer’s and Dementia Translational Research 2017

Randomized, double-blind trial in 34 geriatric patients aged 64 to 92. All completed a multimodal training program over 5 to 7 weeks, plus either hypoxia-hyperoxia or room air. Dementia screening plus 16.7 versus minus 0.39 percent, clock-drawing test plus 10.7 versus minus 8 percent, walking distance plus 24.1 versus plus 10.8 percent. Small sample, surrogate endpoints.

Bayer et al., BMC Geriatrics 2019

Same study population, different endpoints: mobility and self-rated health. Here no difference was found between the groups, neither in the mobility test nor in standing up and walking, in the activities of daily living index or on the health scale. The negative finding stands alongside the positive one from the same group.

Mandel et al., Biomedicines 2023

Randomized trial in 120 patients undergoing bypass surgery with a heart-lung machine. One group received hypoxic-hyperoxic preconditioning beforehand, with oxygen reduction determined individually via the anaerobic threshold. Postoperative complications occurred in 23.3 versus 41.1 percent; markers of the inner vessel lining remained more stable. A preconditioning in hospital, not a longevity protocol.

Where the data stop

For healthy, fit people these data do not exist. The studies are small, short and almost all conducted in sick or very old people, with functional tests and blood markers as endpoints, not with hard endpoints such as disease cases or mortality. A 2022 systematic review of 38 papers puts it clearly: no study shows that intermittent hypoxia extends life expectancy in humans. Even within one study, results diverge, as the geriatric paper shows, in which cognition and walking distance improved but mobility and self-rated health did not. The mitochondria story is a hypothesis from cell and animal data. And the largest study on exhaustion was not randomized and not blinded, with an effort-dependent endpoint.

Status, approval and legal

The procedure is a treatment, not a medicine; there is no approval under medicines law. It is offered in practices, rehab facilities and longevity centers as a self-pay service, and it is not recognized as a service covered by statutory health insurance. The devices are medical devices whose use requires instruction and monitoring of oxygen saturation. This applies all the more to home devices: without saturation monitoring, uncontrolled oxygen deprivation is not a DIY project. Anyone competing in regulated sport should check whether hypoxia procedures are permitted via the current Prohibited List and their own federation.

Safety

In the studies, the procedure was well tolerated, even in patients up to 92 years old. Occasionally mild headaches, tiredness or dizziness are reported at the beginning. The decisive safety factor is the control: oxygen saturation is monitored continuously, and the device adjusts accordingly; without this monitoring, the basis falls away. The procedure is not suitable after a recent heart attack, in unstable heart disease, uncontrolled high blood pressure, severe lung disease, sickle cell anemia, acute infections and during pregnancy. In long COVID, asthma and other pre-existing conditions, the control belongs in experienced hands, and existing treatments are not replaced by it. Anyone who experiences shortness of breath, chest pain or a racing heart during exertion should be examined by a doctor beforehand, not put in a device.

Sources

• Doehner et al., J Cachexia Sarcopenia Muscle 2024 — controlled pilot study on IHHT in long COVID, 145 patients
• Methodological critique of the long COVID study, J Cachexia Sarcopenia Muscle 2025
• Authors’ reply, J Cachexia Sarcopenia Muscle 2025
• Bayer et al., Alzheimers Dement (N Y) 2017 — randomized, double-blind trial on cognition and walking distance in geriatric patients
• Bayer et al., BMC Geriatrics 2019 — same population, no effect on mobility and self-rated health
• Serebrovska et al., Int J Mol Sci 2019 — pilot study in mild cognitive impairment, 21 participants (retracted 2024)
• Bestavashvili et al., Biomedicines 2022 — randomized trial in metabolic syndrome, 65 patients
• Mandel et al., Biomedicines 2023 — randomized trial on hypoxic-hyperoxic preconditioning before bypass surgery, 120 patients
• Tessema et al., Front Aging Neurosci 2022 — systematic review of 38 papers on intermittent hypoxia and aging
• Nobel Prize in Physiology or Medicine 2019 — discovery of cellular oxygen sensing